The Settlers of Catan is one of those games that never quite goes away. True, we go through phases of playing it frequently, and we equally go through phases of not playing it at all, but it's one of those board games, like Mah Jong (BMJA rules, preferably), which never goes away entirely from my life. I think we've been playing it for ten years now. What strikes me, however, is that the way that we play is probably quite different from the way other people play the game - and indeed, that part of the beauty of the game design is that it so readily lends itself to different ways of playing with just a slight tweak to the rules.
The rules as they are written, the spirit of Klaus Teuber's original game, is a highly agonistic, competitive game. Resources are short relative to the number of players (in four player games), and so there is a cut-throat competition to expand or die. This suits the playing style of a lot of different players, and with the potential to be played strategically or tactically, the game has the potential to appeal to anyone for whom Type 1 or Type 2 play is enjoyable.
However, a few trivial changes, and the game suddenly becomes a less hostile, more amiable play experience, suitable for Type 3 play (and either way, because it is a board game and therefore inherently participatory, there is potential for Type 4 play - although we find our modified form is generally more welcoming to a wider variety of players).
The key differences to the way we play are two meta-rules. The first of these is a common house rule, 'burning', which allows you to take back roads or settlements at the end of the chain for a refund of half their resource value. Thus, dead roads can be 'moved' at the cost of half of the resources. I picked this house rule off the internet near the beginning of my time playing the game, and it has stuck. It lets off some of the agonistic pressure, as you can make changes after the fact, and it eases the stress of first road placement, because you don't have to worry about 'dead' roads.
Plus, the burning rule increases the availability of resources, thus making the game play faster. Anything you can do to increase the income and circulation of resources speeds up the game pace, and the second house rule we play with is to eliminate or alter the role of the robber. In the standard game, the robber shuts down the production of the hex it is placed on - it's inherently a competitive (agonistic) element. Change this element, and you produce a friendlier "more Type 3 Wanderer" game, plus you can increase resource production thus making the game flow more easily.
The classic meta-rule we have preferred is to play with the Market, which replaces the Robber. When a 7 is rolled, the player who rolled it gets to trade one card with a pool of five known as the Market. At the start, this is one of each resource. In the early game, the wood and brick often go (depending on the layout of the map), in the later game, sheep, wheat and ore get taken. Either way, trading with the market tends to help most of the time. And because it replaces the Robber (which blocks production), resource acquisition remains evenly paced.
Last night we experimented with the placement of the robber adding one extra resource production to the hex it was placed upon, when that hex's number is rolled. This worked fine, for the most part, but it was still disappointing to roll a 7, because what you want in this game is more resources to build more things, and a 7 means no-one gets any new resources.
I think next time we'll play, we'll make it such that on a 7 you can place the Robber wherever you wish, and when a 7 is next rolled, the hex the Robber is on produces. (Then, you can move the Robber elsewhere, if you wish). Less of a bandit, more of a Migrant Worker, perhaps - when you place it, you are choosing a future payout. "Come to my country, Migrant Worker, and farm sheep for me."
It speaks highly of the elegance of the core design of this game that it is so resilient to change. That we can play it for ten years, and although the core mechanics have remained unchanged, we continue to enjoy tinkering with the meta-rules and so forth to change the way the game plays. In part, this is because the core mechanics are so perfectly abstracted - they contain only what is needed to support play, nothing more and nothing less. This is the epitome of elegant design; it is also what we call tight design - all included elements support the core concept. (It should be noted that the expansions do allow for greater variety, and we are fond of Seafarers of Catan, but the basic game is still perhaps the most elegant).
The way our group plays the game is a strange semi-co-operative, semi-competitive game. Each player is building their own independent economy, but there is some considerable interaction with the other players - both in terms of trade, trade pacts ("I'll help you build roads to that port, in return for free use of it on your turns"), agreeing amicable solutions to land disputes ("I'll build away from your coastal road if you let me settle on that mountain") - and the occasional territorial 'war' when someone decides to act against the common interest. It is not the standard highly competitive game - and I think that these days I would struggle to enjoy to play the game in that way. But the way we play works for a highly diverse set of play styles, and I have not yet found a player who hasn't been able to enjoy it this way (even if the staunchly Type 1 players would greatly prefer the outright bloodshed of the conventional rules).
It is a testament to just how good the core design for this game is that it offers such versatility. I welcome and invite comments from other Settlers of Catan players about their house rules, and the nature of the play experience with those rules - especially when those house rules have been used to 'soften' the competitiveness of the experience, thus making the game more welcoming to more diverse players.